INDEX

Explanations

Korean names and locations

oai_token-act-pair · gpt-3.5-turbo

specific names of individuals, places, or entities in media and politics

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.7.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

-0.72

 NIGHT

-0.69

 barn

-0.69

 goat

-0.65

 olive

-0.63

inct

-0.63

 FedEx

-0.62

AFL

-0.62

 Frenzy

-0.62

 Commonwealth

-0.61

POSITIVE LOGITS

jin

1.22

 Sung

1.11

 Jong

1.07

wei

1.07

jong

1.03

 Jung

1.02

 Kong

1.02

Ji

1.01

ongyang

1.01

Mei

1.01

Activations Density 0.079%

Korean names and locations

specific names of individuals, places, or entities in media and politics

No Comments

No Known Activations